The Music Ontology is an OWL vocabulary for describing music-related information on the Semantic Web. It provides a comprehensive framework for representing artists, albums, tracks, performances, compositions, recordings, and musical works, enabling structured data interchange across music information systems. The ontology has been influential in the Linked Data music community, underpinning projects such as DBTune and informing the design of MusicBrainz's Linked Data exports.
Background
The Music Ontology was created in December 2006 by Yves Raimond as part of his PhD research at Queen Mary, University of London's Centre for Digital Music (C4DM). It arose from the need for a shared vocabulary that could describe the complex relationships involved in music production and distribution -- from the act of composition through performance, recording, signal processing, and release on physical or digital media. The ontology was developed collaboratively with contributions from Thomas Gaengler, Frederick Giasson, Kurt Jacobson, George Fazekas, Simon Reinhardt, and Alexandre Passant.
The project drew inspiration from the FRBR (Functional Requirements for Bibliographic Records) model, adapting its Work-Expression-Manifestation-Item hierarchy to the music domain. This alignment allows the ontology to distinguish between a musical work as an abstract creation, a particular performance of that work, a recording of that performance, and a specific release or file containing that recording.
Purpose & Scope
The Music Ontology addresses the description of music across its full lifecycle:
- Musical Works and Compositions -- abstract creative works, arrangements, and their relationships
- Performances and Recording Sessions -- events where music is performed or captured
- Signals -- analog and digital audio signals, including technical properties like sample rate and bit depth
- Releases and Media -- albums, singles, and the physical or digital media they appear on (CD, vinyl, SACD, digital files, streams)
- Artists and Groups -- solo musicians, bands, labels, and corporate bodies
- Classification -- genres, instruments, and instrumentation
The specification defines 54 classes and 153 properties (207 terms total as of version 2.1.5).
Key Classes
| Class | Description |
|---|---|
| MusicalWork | An abstract musical creation |
| Performance | A performance event |
| Recording | A recorded signal from a performance |
| Signal / DigitalSignal / AnalogSignal | Audio signal representations |
| Release | A specific release of a musical manifestation |
| Track | A track on a particular release |
| MusicArtist / MusicGroup / SoloMusicArtist | Agents involved in music |
| Label | A record label |
| Genre | A musical genre classification |
| Medium (CD, Vinyl, DAT, etc.) | Physical or digital media types |
Serializations & Technical Formats
The Music Ontology is published with the namespace URI http://purl.org/ontology/mo/. The ontology is available in RDF/XML and N3 serializations. It builds upon the Event Ontology (http://purl.org/NET/c4dm/event.owl) and the Timeline Ontology, and references FOAF for agent descriptions.
Governance & Maintenance
The Music Ontology is maintained by its original community of authors. Development is hosted on GitHub under the motools organization. A Google Groups mailing list serves as the primary discussion channel. The most recent version (2.1.5) dates to July 2013, and the ontology has been stable since then.
Notable Implementations
- DBTune -- a suite of Linked Data services exposing music data from sources including MusicBrainz, MySpace, Jamendo, and Magnatune using the Music Ontology
- MusicBrainz -- the open music encyclopedia provides RDF exports aligned with Music Ontology terms
- BBC Music -- the BBC's music platform drew on Music Ontology concepts in its Linked Data architecture
- Linked Data music research -- numerous academic projects in music information retrieval have used the ontology as their RDF vocabulary
Related Standards
The Music Ontology is closely related to the Event Ontology and Timeline Ontology (both developed at C4DM), and to FOAF for describing agents. It shares conceptual alignment with the FRBR model, from which its Work/Expression/Manifestation/Item hierarchy derives.